First Reported Nonpeptide AT1 Receptor Agonist (L-162,313) Acts as

Sebastian T. Le Quement , Thomas E. Nielsen and Morten Meldal ... Plouffe, Hélène Beaudry, A. K. Mahalingam, Xiongyu Wu, Berndt Johansson, Mathias H...
0 downloads 0 Views 195KB Size
1536

J. Med. Chem. 2004, 47, 1536-1546

First Reported Nonpeptide AT1 Receptor Agonist (L-162,313) Acts as an AT2 Receptor Agonist in Vivo Yiqian Wan,† Charlotta Wallinder,† Berndt Johansson,‡ Mathias Holm,‡ Mahalingam A. K.,† Xiongyu Wu,† Milad Botros,‡ Anders Karle´n,† Anders Pettersson,‡ Fred Nyberg,§ Lars Fa¨ndriks,‡ Anders Hallberg,† and Mathias Alterman*,† Department of Medicinal Chemistry, BMC, Uppsala University, P.O. Box 574, SE-751 23 Uppsala, Sweden, Department of Gastrosurgical Research, Sahlgrenska Academy at Gothenburg University, SE-413 45 Gothenburg, Sweden, and Department of Biological Research on Drug Dependence, BMC, Uppsala University, P.O. Box 591, SE-751 23 Uppsala, Sweden Received September 12, 2003

In this investigation, it is demonstrated that the first nonpeptide AT1 receptor agonist L-162,313 (1), disclosed in 1994, also acts as an agonist at the AT2 receptor. In anesthetized rats, administration of compound 1 intravenously or locally in the duodenum increased duodenal mucosal alkaline secretion, effects that were sensitive to the selective AT2 receptor antagonist PD-123,319. The data strongly suggest that 1 is an AT2 receptor agonist in vivo. To the best of our knowledge, this substance is the first nonpeptidic low-molecular weight compound with an agonistic effect mediated through the AT2 receptor. Introduction Selective AT1 receptor antagonists block well-known functional effects of the octapeptide angiotensin II (Ang II), such as vasoconstriction, aldosterone release, and cardiovascular growth.1 In 1994, the first nonpeptide Ang II agonist, L-162,313 (1) was disclosed (Figure 1).2 At that time, nonpeptide agonists of peptide receptors were rare, being confined largely to agonists of opioid receptors. Notably, removal of a methyl group from L-162,782, an agonist closely structurally related to the thiophene derivative 1, provided L-162,389 which was found to act as an antagonist in vivo.3 Thus, a subtle molecular alteration was shown to determine the agonist/ antagonist properties of these ligands. The three compounds described above were all essentially equipotent as AT1/AT2 receptor binding ligands, but by replacing the alkyl group with a m-methoxybenzyl group, an agonist, L-163,491, with an approximately 70fold selectivity for the AT1 receptor subtype, was obtained.4 Ang II exhibits a similar binding affinity to the AT1 and AT2 receptors, sharing a sequence homology of only 32-34% at the amino acid level in rat.5 We felt encouraged to answer the question of whether 1 also would exert agonistic properties toward the AT2 receptor.5,6 Recently, we reported an AT2 receptor-mediated Ang II stimulation of duodenal mucosal alkaline secretion in the rat,7 an effect that could be blocked by the AT2 selective antagonist PD-123,3198 (35) but not by the AT1 receptor antagonist losartan.9 Prior to selecting the compounds for studies in this animal model, we decided to (a) assess the impact on AT1/AT2 receptor affinity caused by minor alterations of the sulfonylcarbamate * Corresponding author. Tel: +46-18-4714905. Fax: +46-184714474. E-mail: [email protected]. † Department of Medicinal Chemistry, Uppsala University. ‡ Sahlgrenska Academy at Gothenburg University. § Department of Biological Research on Drug Dependence, Uppsala University.

Figure 1.

part of 1 (Series A, Figure 2) and (b) to determine the receptor selectivities after retaining the lower part of 1 but replacing the bicyclic imidazopyridine ring system with substituted quinazolinones, as the latter structure is found in a large number of selective AT2 receptor antagonists (Series B, Figure 2).10 We herein report that the AT1 receptor agonist 1, the compound with the most favorable AT2/AT1 affinity ratio in the A series acts as an AT2 receptor agonist in the animal model. In contrast, compound 12 in series B, with a 40-fold greater affinity for the AT2 receptor versus the AT1 receptor, exerts no agonistic effects in this in vivo model. Chemistry The thiopheneboronic acid 2, a key intermediate for the synthesis of the compounds in both series, was prepared essentially as described by Kevin et al.11 Thus, thiophene-2-sulfonyl chloride was first converted to the N-tert-butylsulfonamide. Subsequent alkylation followed

10.1021/jm031031i CCC: $27.50 © 2004 American Chemical Society Published on Web 02/14/2004

AT2 Receptor Agonist in Vivo

Journal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1537

Figure 2.

Scheme 1a

a Reagents and conditions: (a) Pd(PPh ) , NaOH (aq), ethanol/toluene; (b) TFA; (c) alkyl chloroformate or acyl chloride, pyrrolidinopy3 4 ridine, pyridine.

by selective 3-lithiation/boration delivered 2. A Suzuki coupling of 3-(4-bromobenzyl)-2-ethyl-5,7-dimethyl-3Himidazol[4,5-b]pyridine12 with the boronic acid 211 provided 3 in good yield. Deprotection by TFA, to give the primary sulfonamide followed by reactions with alkyl chloroformates and acyl chlorides, afforded 12, 4-6, and 7-9, respectively (Scheme 1). A Suzuki coupling of 2 with 3-(4-bromobenzyl)-6-nitro2-propyl-3H-quinazolin-4-one yielded 10 in high yield. The ethyl group most often found in the 2-position of the quinazolin-4-one scaffold of AT2 receptor antagonists was replaced with propyl since Glinka et al. had observed that the propyl group in a related series of compounds gave higher AT2 receptor affinities when combined with the sulfonamide based carboxylic acid bioisostere.10c The compounds in series B were prepared as outlined in Scheme 2. Deprotection of 10 and reaction with butyl chloroformate gave 11 in a good yield. To obtain the derivatives 12-21, the nitro group of 10 was first reduced using ammonium formate and palladium on charcoal to provide 22. Acylation of the amine function of 22 with acetyl chloride, benzoyl chloride, and thienoyl chloride respectively afforded compounds 23-25 which were debutylated and then reacted with butyl chloroformate to yield the secondary amides 12-14. The tertiary amides 15-21 were prepared by reductive alkylation using acetaldehyde or benzaldehyde with triacetoxyborohydride as reducing agent to give 26 and 27, respectively. These secondary amines were thereafter acylated to afford 28-34 and subsequently deprotected and treated with butyl chloroformate to deliver the desired tertiary amides. Binding Assays. Compounds 1, 4-9, 11-21 were evaluated in radioligand-binding assays by displacement of [125I]Ang II from AT1 receptors in rat liver membranes and from AT2 receptors in pig uterus membranes in essence as described previously (Table 1).13 The natural substrate Ang II, the selective AT1

Table 1. Series A

a K values are an average from three determinations. Standard i deviations are less than 15% in all cases.

receptor antagonist losartan,9 and the selective AT2 receptor antagonist 358 were used as reference substances. In Vivo Assays. The in vivo experiments were performed on anaesthetized nonfasted male SpragueDawley rats. A femoral artery and one or two veins were catheterized for subsequent blood pressure measurements and drug infusions, respectively. Duodenal mu-

1538

Journal of Medicinal Chemistry, 2004, Vol. 47, No. 6

Wan et al.

Scheme 2a

a Reagents and conditions: (a) Pd(PPh ) , NaOH (aq), ethanol/toluene; (b) TFA; (c) butyl chloroformate, pyrrolidinopyridine, pyridine; 3 4 (d) Pd/C, HCO2NH4, MeOH; (e) acyl chloride, DIEA, CH2Cl2; (f) acetaldehyde or benzaldehyde, NaB(OAc)3H, AcOH, CH2ClCH2Cl.

cosal alkaline secretion (HCO3- secretion) was measured by a pH-stat titration technique.14 Alkaline secretion to the perfusate was continuously titrated to pH 7.4 with 0.02 M HCl controlled by a pH-stat device. Results and Discussion The results obtained in the AT1 and AT2 receptor binding assays are presented in Tables 1 and 2. The Ki values for compound 1 were determined as 3.9 nM for the AT1 and 2.8 nM for the AT2 receptors (lit. data IC50: 1.1 nM and 2.0 nM, respectively2b). As apparent from Table 1 all variations of the butyloxy group investigated were deleterious with respect to AT2 receptor affinity, although low affinity was encountered by shortening of the side-chain by removal of the oxygen atom (compound 9). Notably, the ethyloxy derivative 4 exhibited no detectable AT2 receptor affinity. The lack of AT2 receptor affinity exhibited by 7 is also somewhat remarkable considering the previous results by Mantlo et al.,15 who in their series of compounds observed a higher selectivity for the AT2 receptor with compounds bearing the cyclopentyl side-chain. With regard to AT1 receptor affinity, the impact of alterations to the butyloxy side-chain was considerably less pronounced. All compounds bound in varying degree to the AT1 receptor. Thus, the effect caused by alterations of the butyloxycarbonylsulfonamide portion of 1 seemed to resemble the outcome of variations of the isobutyl group in the

5-position, as reported by Kevin et al.,11a in that even the small structural modifications performed tended to reduce AT2 receptor affinity while retaining AT1 receptor affinity. The most potent AT2 receptor ligand in series A, i.e. 1 (not receptor subtype selective but a proven partial AT1 receptor agonist) was investigated in the rat in vivo. Duodenal mucosal alkaline secretion in rats has been shown previously to be inhibited by AT1 receptor activation and increased after AT2 receptor stimulation.7,16 As shown in Figure 3, intravenous administration of 1 (bolus 0.3 mg/kg plus 30 µg/kg×h) increases mean arterial pressure by approximately 10 mmHg. This pressor effect was reversed, with a markedly lowered arterial pressure by addition of the AT1-receptor antagonist losartan (10 mg kg-1 iv bolus) indicating that 1 activates AT1 receptor-mediated vasoconstriction. Interestingly, the compound alone also increased the duodenal mucosal alkaline secretion moderately. The addition of losartan markedly increased this secretory stimulation (Figure 3). When losartan was combined with the AT2 receptor antagonist 35 this effect was absent (data not shown, n ) 3). Although a proper doseresponse-curve was not performed with regard to intravenous administration, the effect on mucosal alkaline secretion by compound 1 at the presently used infusion rate can be considered of the same order of magnitude as previously reported for the peptide compounds Ang

AT2 Receptor Agonist in Vivo

Journal of Medicinal Chemistry, 2004, Vol. 47, No. 6 1539

Table 2. Series B

Figure 3. Anesthetized Sprague-Dawley rats (n ) 5). Effects of intravenous administration of 1 (bolus 0.3 mg/kg plus 30 µg/kg × h) and later addition of the AT1 receptor antagonist losartan (10 mg/kg bolus iv) on duodenal mucosal alkaline secretion and mean arterial pressure.

Figure 4. Anesthetized Sprague-Dawley rats. Effects of local (intra duodenal) administration of 1 at consecutively increased concentrations in 30 min periods in absence (n ) 5) and in the presence (n ) 5) of 35 (0,1 mM). Values represent means of the last 15 periods of each concentration.

II and CGP112A.7 The results support the view that compound 1 is an unselective agonist at both the AT1 and AT2 receptors. However, to investigate the AT2 agonistic properties of 1 in greater depth, topical administration was employed by administering the compound in the duodenal intraluminal perfusate. The rationale behind this mode of administration was that AT2 receptors, but not AT1 receptors, are localized to the secreting epithelium7 making simultaneous administration of an AT1 receptor antagonist unnecessary. It was observed that 1 given in the perfusate raised the alkaline secretion in a concentration dependent manner (Figure 4). Simultaneous presence of 35 (0.1 mM) significantly inhibited this effect indicating that an AT2 receptor-mediated effect was occurring. Such local intraluminal administration of drugs did not influence arterial pressure (data not shown). Taken together these in vivo tests strongly suggest that 1 is a dual AT1/AT2 receptor agonist. Since the data in series A and previous findings3 suggested that the 2-(N-butyloxycarbonyl)sulfonamide and 5-isobutyl were compulsory for fair AT2 receptor affinity to be achieved, we turned our attention to modifications of the bicyclic upper part of the molecule in the hope of improving the AT2/AT1 ratio. We hypothesized that the lower part of 1 might be pivotal for agonism while a proper upper heterocyclic moiety could

conceivably provide high AT2 receptor selectivity. We were particularly attracted by a series of AT2 receptor ligands of the biaryl type reported by Glinka et al.10 Some of these compounds exhibited an impressive AT2/ AT1 selectivity (e.g. for one of the derivatives, AT1: Ki >3000 nM and AT2: Ki ) 0.06 nM). The compounds were comprised of a quinazolin-4-one scaffold substituted in the 6-position with, for example, a variety of acylamido groups. Partly guided by the results of Glinka et al., the compounds in Table 2 were designed and synthesized. From Table 2 it is apparent that the substituent in the 6-position of the quinazolinones can significantly affect the biological activity. While all amides (12-21) bind to the AT1 receptor, albeit weakly, minor structural variations could have a dramatic influence on the affinity to the AT2 receptor. The thiophene and the corresponding benzene rings in 1 and L-162,782, respectively, act as true bioisosteres, but these aromatic nuclei are not necessarily interchangeable when constituting a part of the 6-substituent of the quinazolinone moiety. A comparison of the tertiary ethyl amides 16 and 17 suggests that interchanging the two aromatic systems has no large impact on the binding affinity to the AT2 receptor. However, substitutions of the ethyl group of the thiophene derivative 17 with either hydrogen (cf. 17 and 14) or a benzyl group (cf. 17 and 21) are deleterious for AT2 receptor affinity. However, when the benzene analogue 16 was subjected to the same modi-

1540

Journal of Medicinal Chemistry, 2004, Vol. 47, No. 6

Wan et al.

fications, surprisingly, the affinity was largely retained (cf. 16, 13, and 20). These results are difficult to rationalize but demonstrate a complex structure-activity relationship of the compounds in series B. Thus, as in series A where the AT2 affinity drops significantly after minor structural changes to the oxybutyl chain of the prototype compound 1 or of the isobutyl group, as previously shown,11a a similar response is encountered after the “bioisosteric” phenyl to thienyl exchange, conducted in the series B. One of the most potent and AT2 receptor selective ligands in series B, the secondary amide 12 (the ligand with the most favorable solubility properties in the series), was selected for in vivo studies in rat. Compound 12 was not soluble in ethanol unlike 1. Ethanol exerts only marginal effects on the studied secretory variable at the used concentration (